Spatio-temporal hierarchical cluster analysis of mining-induced seismicity in coal mines using Ward's minimum variance method

Abstract

Mining-induced seismicity is one of the most dangerous physical phenomenon's accompanying coal and ore extraction in underground mines and often resulting in rock burst. In many cases, this seismicity can be grouped together, forming spatial clusters of events related to some structures in mines such as faults or pillars. However, as shown in this paper, the usual seismicity in coal mines is mainly related to current mining process and there are no distinct spatial or temporal groups of seismic events, but rather they form continuous cloud of events related to mining activity. Therefore, we introduce new spatio-temporal metric together with Ward's minimum variance method to obtain hierarchical clustering analysis of mining-induced seismicity. Constructing metric in this way made it possible to assess seismic hazard concurrently in space and time. The clustering method is utilized to a case study of mining-induced seismicity from coal mine with high seismic hazard and where high energy mainshock occurred. In each cluster, magnitude distribution of seismic activity is calculated, showing in some cases, a significant departure from Gutenberg-Richter distribution of seismic events, which can be important in seismic hazard analysis. As a measure of accuracy of each cluster locations we have introduced bootstrapping procedure. We have shown that location errors of seismic activity affect each cluster center location of the order of a few tens of meters. The presented cluster analysis of mining seismicity provides new measures to determine group of seismic sources and high stressed areas in the rock mass in mines allowing for determination of seismic and rockburst hazard.